Add support for more advanced backend checks, e.g. haproxy agent-check

[Daniel Failing]

Since haproxy 2.2 there exists support for more advanced health checks. [1]
This will also allow the backend/member to tell the haproxy to reduce the load or put a node in draining, rather than only UP/DOWN checks.

For supporting it, there probably has to be another healthmonitor type.
The agent-check, agent-address, agent-port, agent-inter should be able to be inferred from the current values already. agent-send is something not yet in the API, but could probably be extended at a later point.

As an additional step, the operating states of the members could be expanded to allow for more states like DRAIN/MAINT
Otherwise a translation from those states to ERROR could work.

[1] https://www.haproxy.com/documentation/haproxy-configuration-tutorials/service-reliability/health-checks/#agent-checks